Involvement of circadian clock protein PER2 in controlling sleep deprivation induced HMGB1 up-regulation by targeting p300 in the cortex

Abstract

Lack of sleep is a common problem in current society, which can induce various brain dysfunctions. Neuroinflammation is a typical reaction caused by sleep deficit and is considered as a common basis for various neurological disorders and cognitive impairments, but the related mechanisms have not been fully clarified. The circadian clock protein plays a critical role in maintaining physiological homeostasis, including sleep/wake cycles. Circadian disorders induced by sleep deficit might contribute to the development of neuroinflammation. In the current study, we observed that sleep deprivation (SD) induced elevated expression of High-mobility group box 1 (HMGB1), one of the most important mediators of neuroinflammation, in the cortical microglia and cerebrospinal fluids. Moreover, acetylation-dependent nuclear export of HMGB1 was involved in up-regulation and secretion of HMGB1 after sleep deprivation. Further studies indicated that sleep deprivation induced an increase in the expression of acetyltransferase p300 and a decrease in the expression of deacetylase SIRT1, which synergistically enhanced the acetylation level of HMGB1 in the cortical microglial cells, thereby triggered the nuclear export and secretion of HMGB1. Most importantly, circadian clock protein PER2 constitutively interacted with p300 and inhibited its expression in the microglial cells, which can be interrupted by PER2 downregulation upon sleep deprivation, leading to the increased expression of p300 and acetylation and secretion of HMGB1. The truncated PER2 mutant without p300 binding ability lost its ability to regulate p300 expression, indicating that PER2 functioned as a co-suppressor of p300 in regulating acetylation and expression of HMGB1. Taken together, data in this study reveal a new mechanism by which PER2 is involved in controlling HMGB1 dependent neuroinflammation induced by sleep deprivation. Maintaining PER2 levels or blocking HMGB1 acetylation in the cortex might be prospective for preventing sleep deprivation-induced neuroinflammation and the related adverse reactions in the brain.

Keywords: HMGB1; Neuroinflammation; PER2; Sleep deprivation; p300.

Fig. 1

Sleep deprivation induced HMGB1 upregulation in the microglia of cortex and cerebrospinal fluids (CSF). 6-8-week-old male Sprague-Dawley rats were subjected to continuous sleep deprivation for 72 h, and then (A) HMGB1 levels in CSF before and after sleep deprivation were determined by ELISA. (N = 6) (Band D) HMGB1 levels in the cortex (B) and hippocampus (D) before and after sleep deprivation were detected by western blot assay. (N = 3) (C and E) Relative expression and quantification of the proteins in Fig. 1B (C) and 1D (E) were shown, respectively. (F–H) The expressions and quantification of HMGB1 in the cortical microglia (IBA-1-positive cells, F), astrocyte (GFAP-positive cells, G) and neuron (NeuN-positive cells, H) before and after sleep deprivation were detected by co-staining of HMGB1 with each cell marker as indicated in the immunofluorescence assay. Data are expressed as the mean ± SD. Scale bar: 25 μm; NS: no significance, * p < 0.05, **p < 0.01.

Fig. 2

Sleep deprivation induced acetylation-dependent nuclear export of HMGB1 in the cortex. (A) Rats were subjected to sleep deprivation for 72 h, and then the expression levels of cortical HMGB1 mRNA before and after sleep deprivation were detected with RT-PCR assay. (N = 3) (B)Relative expression and quantification of the mRNA in Fig. 2A were shown. (C) Rats were subjected to sleep deprivation for 72 h, and then cortex tissue lysate was subjected to immunoprecipitation with the anti-HMGB1 antibody followed by blotting with anti-acetyl-lysine antibody. (N = 3) (D) Rats were subjected to sleep deprivation for 72 h, and then the cytoplasmic/nuclear distribution of HMGB1 in the cortex tissue lysate before and after sleep deprivation were detected. (N = 3) (E) Relative expression and quantification of the proteins in Fig. 2D were shown. Data are expressed as the mean ± SD. NS: no significance, * p < 0.05, **p < 0.01.

Fig. 3

p300 and SIRT1 regulated HMGB1 acetylation in the microglial cells. (A) Rats were subjected to sleep deprivation for 72 h, and then the expression levels of p300, CBP and SIRT1 in the cortex before and after sleep deprivation were detected. (N = 3) (B–D) Relative expression and quantification of the proteins in Fig. 3A were shown. (E and H) BV2 (E) or C8D1A (H) cells were transfected with p300 siRNA or its control siRNA, and then the expression levels of p300 and HMGB1 were detected at 36 h after transfection. (F and I) BV2 (F) or C8D1A (I) cells were treated with SIRT1 inhibitor EX527 or its solvent DMSO, and then the ratio of cellular NAD+/NADH was detected at 24 h later. (G and J) BV2 (G) or C8D1A (J) cells were treated as described in Fig. 3F and I, and then the expression levels of SIRT1 and HMGB1 were determined at 36 h later. Data are expressed as the mean ± SD. siRNA Transfection and inhibitor treatment in BV2 and C8D1A cells were repeated 3 times independently, the representative bands and their relative levels to the control bands are shown. NS: no significance, * p < 0.05, **p < 0.01.

Fig. 4

Impairment of the circadian clock protein PER2 expression resulted in p300-dependent HMGB1 upregulation in the microglia upon sleep deprivation. (A) Rats were subjected to sleep deprivation for 72 h, and then the expression levels of the circadian clock proteins (CLCOK, BMAL1, PER2, CRY1, CRY2) in the cortex before and after sleep deprivation were detected. (N = 3) (B) Relative expression and quantification of the proteins in Fig. 4A were shown. (C) Rats were subjected to sleep deprivation for 72 h, and then the expression levels of PER2 in the hippocampus before and after sleep deprivation were detected. (N = 3) (D) Relative expression and quantification of PER2 in Fig. 4C were shown. (E and F) BV2 (E) or C8D1A (F) cells were transfected with PER2 siRNA or its control siRNA, and then the expression levels of PER2, p300, SIRT1 and HMGB1 were determined at 36 h later. Data are expressed as the mean ± SD. siRNA and plasmids transfection in BV2 cells were repeated 3 times independently, the representative bands and their relative levels to the control bands are shown. NS: no significance, * p < 0.05.

Fig. 5

PER2 acted as a co-suppressor of p300 in microglial cells. (A) BV2 cells were transfected with FLAG-PER2 expression plasmid or its control vector, and then the whole cell lysate were immunorepcipitated with anti-FLAG antibody followed by blotting with anti-p300 antibody to detect the potential interaction of PER2 and p300. (B) Design of the truncated PER2 mutants. (C) BV2 cells were transfected with WT-PER2 or the truncated PER2 mutants, and then the whole cell lysate were immunorepcipitated with anti-FLAG antibody followed by blotting with anti-p300 antibody to detect p300 binding domain within PER2. (D) BV2 cells were transfected with PER2 siRNA or its control siRNA, followed by reconstitution with WT-PER2 or the truncated PER2 mutants. Then the expression levels of p300 were detected at 36 h later.

Fig. 6

Schematic presentation of working model in this study. PER2 constitutively interacted with p300 and inhibit p300 expressions in the cortical microglia under steady state. Sleep deprivation induced PER2 downregulation and interruption of PER2/p300 interaction. Then the increased expression of p300 triggered HMGB1 acetylation and secretion, leading to the elevation of HMGB1 in the cortex and CSF and the resultant neuroinflammation.